Literature Review on Teacher Education in the 21st Century

5 PROFESSIONALISM AND PUPIL OUTCOMES

Key points summary It is widely agreed that teaching quality is linked to pupil outcomes and it is often suggested that teacher education contributes to the development of teacher quality. However, teaching quality is difficult to define and the concept of pupil outcomes related to teacher quality is contested. These and other factors create considerable methodological challenges for researching these matters. There is very little published work on 'wider' pupil outcomes. There is conflicting evidence (mainly from the USA) about the impact of higher level teaching qualifications on pupil attainment.

Introduction

5.1 This section of the report addresses the third objective of the literature review and explores the relationships between enhanced professionalism and pupil outcomes. It is widely accepted that teaching quality is a key determinant of successful schooling, as suggested by the McKinsey Report (2007) and by the European Commission:

Research shows that teacher quality is significantly and positively correlated with pupil attainment and that it is the most important within-school aspect explaining student performance…Furthermore, other studies have found positive relationships between in-service teacher training and student achievement. (European Commission, 2007: 3)

5.2 Notably, the European Commission report cites an Israeli study that suggests that teacher training 'may provide a less costly means of increasing test scores than reducing class size or adding school hours' (ibid.)

5.3 The definition of teaching quality and its measurement remain problematic and the evidence reviewed is far more equivocal than the EC report suggests. Furthermore, there is little research exploring the links between teacher education and pupil outcomes. The research that does exist is limited in its scope. Some definitions of teaching quality may derive simply from measuring pupil outcomes and this leads to a somewhat circular definition of both terms - you judge good teaching by pupil outcomes, therefore the best teachers are the ones where pupils have the best outcomes

5.4 The term 'pupil outcomes' is hotly contested. Given the contemporary global influence of international benchmarking studies such as the Programme for International Student Assessment, the Trends in International Mathematics and Science Study and the Progress in International Reading Literacy Study, there is a tendency to define pupil outcomes simply in terms of attainment results in standardised tests. However, such results are but one part of the outcomes that may reasonably be expected of education. In the Scottish context, the achievement and development of the four 'capacities' of Curriculum for Excellence - successful learners, confident individuals, responsible citizens, effective contributors - will not be fully revealed by such data. This review has identified very little work that links teacher education directly to the 'wider personal development of young people' referred to in the aim of the study.

5.5 However, if it is accepted that teachers do vary in their quality and that the outcomes achieved by pupils do matter, then it is important to attempt to clarify these concepts so that meaningful investigation and discussion can inform policy, not least on forms of teacher education that will lead to improvement in teaching and outcomes.

Research linking teaching education and pupil outcomes

5.6 The literature search carried out for this review has revealed remarkably few studies that examine the influence of teacher education on pupil outcomes. The great majority of efforts to undertake such work have been carried out in the USA and this work is summarised here - as previously noted in Section 1 the U.S. education system has many differences to the Scottish system and this should be noted when considering the transferability of research findings. One of the reasons for the paucity of such work is undoubtedly the methodological complexity of devising studies where linkages between teacher education and pupil outcomes can be isolated from other factors.

5.7 Borko et al (2008) suggest that research on the 'Effects of Teacher Education' has its methodological roots in natural sciences and is concerned to identify cause and effect. They suggest it has become more prevalent in the wake of the No Child Left Behind agenda in the USA and the currency of research that is designed to identify 'what works'. However, although believed by policymakers to have useful predictive value Borko et al suggest that '[such] studies cannot always account for why something works or fails to work in particular contexts' (Borko et al, 2008: 1024).

5.8 In a review of work in the USA that uses standardised student achievement test scores as the outcome measure, Goe (2007:2) points out that even such a uniform approach has difficulties, including (emphases in original):

• Standardised achievement tests were intended to measure student achievement and were not designed to measure teacher quality.
• It is difficult to sort out teacher effects ( i.e. the contribution of teachers) from classroom effects ( i.e. the contribution of peers, textbooks, materials, curriculum, classroom climate, and other factors).
• It is difficult to obtain linked student-teacher data that make it possible to connect specific teachers to student achievement test scores.

5.9 Goe (2007) uses two 'input measures', teacher qualifications and teacher characteristics, a process measure of teacher practices and an outcome measure, teacher effectiveness. She concludes that the findings of the reviewed studies are inconsistent or of no practical significance even when the findings are statistically significant.

5.10 In 2008, Cochran-Smith and Fries published 'a review of reviews' of research on teacher education from the 1920s to 2005. They report that Ballou and Podgursky (2000a, b) found that the evidence that teacher effectiveness is enhanced by advanced degrees earned in Schools of Education is weak. However, Darling-Hammond (2000) re-examined the same material and concluded that Ballou and Podgursky had misrepresented most of the existing evidence base in order to argue that teacher education makes no difference to teacher performance or student learning (Cochran-Smith and Fries, 2008: 1062). The Abell Foundation (2001) reviewed 150 papers published over 50 years and came to the conclusion that:

The academic research attempting to link teacher certification with student achievement is astonishingly deficient; certification is incapable of providing any insight into an individual's curiosity, affinity for children, and/or instructional skills. (ibid: 1062-1063)

5.11 Cochran-Smith and Fries also report that Wilson et al (2001) reviewed 57 studies over the period 1985-2000 and showed a positive connection between teacher preparation in their subject matter and their performance and impact in the classroom (ibid.: 1063). They also refer back to Report of the American Educational Research Association Panel on Research and Teacher Education edited by Cochran-Smith and Zeichner (2005), where the editors found that:

Research on certification is limited, but the weight of evidence generally favors certification over non-certification or under-certification, as measured by student achievement. 42 states require teacher testing, but there is little evidence these have predictive validity. Studies on the impact of accreditation are almost non existent. (ibid: 1066)

5.12 A recent report from the Committee on the Study of Teacher Preparation Programs in the United States reaches similar conclusions (National Research Council, 2010:22). They point to the methodological and conceptual difficulties in establishing causal links between teacher preparation and pupil outcomes:

• There are no well-formed theories that link teacher preparation to student outcomes.
• The complex nature of schooling children makes it difficult to identify empirically the role of teacher preparation among the many intertwined influences on student outcomes.
• The use of strict experimental design principles can be problematic in some educational settings…it is difficult to control for all the important factors that are likely to influence student outcomes.

5.13 The Committee ( NRC, 2010) also considered the wider issues of quality control in teacher education (see also Section 6 below) and pointed out that in the USA, as elsewhere, there are procedures for ensuring quality at individual and at institutional level. They point out that there are many difficulties in 'teacher tests', not least being confident that the items measured are significant in teacher performance. Referring to institutional quality, they review the application - at state level in the USA - of Standards and point out:

The standards that do exist are not based on research that demonstrates links between particular standards and improved outcomes for students taught by teachers who were educated in a particular way because such evidence is not available. …. We note that teacher education is hardly alone in lacking data that directly link components of professional preparation to the outcomes for those who receive the professionals' services⁹. ( NRC, 2010: 159)

5.14 A study in North Carolina over the years 1994-2004 reported that elementary school students in Grades 2 to 5 fared better in math and reading tests when they had been taught by teachers with National Board Certification ( NBCTs) (Clotfelter et al, 2007). Similarly, a small-scale study by Sato et al (2008) reported higher quality assessment practices among NBCTs. However, such findings have been challenged by Rouse (2008) who did not find a significant relationship between board certification and pupil attainment in his quasi-experimental study of 54 teachers in North Carolina. In other studies, Gimbert et al (2007) have attempted to relate different models of teacher preparation to student attainment but could find no correlation whilst Lustick and Sykes (2006) found significant achievements in teacher learning through Board certification, but did not consider student outcomes. However, in a survey of NBCTs due to renew their certification (after ten years), 98% reported that National Board Certification had positively influenced their careers and 92% reported that National Board Certification had positively influenced their students' learning (Petty et al, 2007).

5.15 In the UK, work undertaken by the Centre for Market and Public Organisation at Bristol University (Slater et al, 2009) has examined the effect of individual teachers on pupil outcomes. Using a definition of teacher quality restricted to impact on student outcomes, Slater et al (2009) use a unique dataset containing the GCSE examination results and Key Stage 3 test results in maths, science and English for 7,305 pupils taught by 740 teachers across 33 schools in England between 1999 and 2002. Pupil records are linked with particular teachers through class lists provided by the schools. The analysis considered subject-specific prior achievement (for previous teacher effects) and observable school effects (intake, resources, selection). According to the findings of this (non peer reviewed) study, teacher characteristics of gender, age, experience and education do not play any statistically significant role in explaining variability in teacher effectiveness. A negative effect was shown only in relation to very low levels of experience. The authors concur with research conducted in the US (Kane et al, 2008) in concluding that teacher characteristics are not reliable indicators of teacher quality.

5.16 One English study of the influence of CPD took a longitudinal approach and did claim that there were detectable changes in teaching style that derived from the professional development experiences (Boyle et al, 2005). Those CPD experiences that were found to be most effective were the longer term ones that included peer observation and sharing of practice. However, a relatively large scale study of CPD in England found that only 24% of schools were attempting to evaluate CPD undertaken by teachers in terms of the influence on pupil attitudes (Goodall et al, 2005).

5.17 Little is known about the relationship between characteristics of different professional development activities and pupil outcomes. The evaluation of the GTCETeacher Learning Academy ( TLA) (Lord et al, 2009) provides some evidence to support a link between enquiry-based learning and positive outcomes for teachers, pupils and schools. The evaluation found evidence of impact on teachers' capacity to reflect on practice and self-evaluation, school policies ( e.g. behaviour management, CPD approaches and pupil involvement), pupils' progress (before and after tests, and using Assessment for Learning approaches). However, the evaluators noted that 'schools did not appear to have a systematic approach to evaluating CPD and its impact' (p.109) and recommended greater use of pre- and post-intervention methodologies by teachers and CPD leaders to support robust assessment of impact on classroom practice and pupil learning.

5.18 A systematic review of induction studies (Totterdell et al, 2004) indicated that the lack of large-scale and longitudinal research in this area prevents the type of investigation that might lead to a sound understanding of the connection between enhanced professionalism and the quality of pupil outcomes.

5.19 A systematic review of subject specialist input into CPD in England did find some evidence of impacts on pupils in the following areas: learning and achievement ( e.g. improved knowledge of scientific concepts, problem solving, mathematical skills, literacy skills, reasoning, and use of ICT) as well as affective development ( e.g. attitudes to learning, motivation and self-esteem) (Cordingley et al, 2007).

5.20 In Scotland, a study of CPD for science teachers by Thurston et al (2008) suggests that it was possible to measure impact in terms of increased pupil attainment by relating this to the changes in classroom practice attributed to CPD. They conclude that CPD can facilitate changes in the professional practice of teachers; however, it must be supported by well structured opportunities allowing teachers to draw support and advice from each other.

5.21 Some writers have positively associated action research partnerships between schools and universities with improving pupil outcomes ( e.g. Costello et al 2000; Slater and Ravid, 2010). Similarly, mentor training and development has been found to lead to improvements in teaching and learning (Strong, 2009). A small-scale survey (95% response rate) by Dallat et al (2000) of 20 teachers following a one-year course in Expert Teaching in Northern Ireland concluded that changes to practice are most likely to occur where teachers : have time to reflect and review their practice; participate in collegial discussions and observations to share practice and encourage professional development; learn in their school context; and undertake longer term professional development.

5.22 As Day et al (2005; 2006a, b) have shown teacher commitment is very important in these matters. They argue that professional development should not be divorced from the need for wider contextual understanding of what enhances teacher commitment to the profession; taking a standards-based approached to professional development may serve to decrease commitment. They suggest that policymakers and school leaders should consider the contexts for professional development which can change practice positively - that is, contexts in which teachers' professional values are acknowledged and built on. If commitment can be sustained across the career phases then problems of teacher retention are less likely to occur. Day and Gu (2007) also suggest there is a particular lack of consideration given to the CPD needs of more experienced teachers.

5.23 In England, Hurd (2008) used inspection evidence to assess whether the presence of student teachers in secondary schools had an effect on pupil attainment. The study considered more than 1200 schools over a three year period.

The number of trainees has no significant effect on school results at A-level or General Certificate of Secondary Education ( GCSE), or on the overall value added between Key Stage 3 and GCSE level. However, at Key Stage 3 level at age 14, while there appears to be a very small depressing effect on achievement in schools with low numbers of trainees, there is a significant positive effect on achievement in schools with larger numbers of trainees. (Hurd, 2008:19)

In a previous article, Hurd et al (2007) found that involvement in ITE appeared to have a positive influence on teachers' professional development but found it difficult to relate this to pupil outcomes.

5.24 In the USA some Professional Development Schools ( PDS) have explicitly sought to bring improvements in pupil outcomes and teacher education together. The Kansas State University PDS Partnership project ¹⁰ offers an example of an initiative with the dual aim of improving pupil outcomes and improving teacher preparation. The project shows significant gains in student achievement and positive outcomes of reform of the teacher education programme (Shroyer et al, 2007). The report authors acknowledge that sustained and intensive work requires investment, support for professional development and change among faculty, school staff and administrators.

"If the ultimate vision for teacher education is to enhance K-12 student learning, then teacher educators in K-12 schools and colleges of education and arts and sciences must perceive themselves as directly responsible for the teaching and learning of K-12 students as well as that of future and practising teachers." (Shroyer et al 2007, p. 223)

5.25 Research also indicates that leadership is crucial in securing improvements in pupil outcomes. The major review of such literature, carried out in New Zealand by Robinson et al (2009) found the most important aspect is that school leaders must be active in areas of teacher learning and development. The leader must be seen to be active to demonstrate to teachers that he/she sees the value of what is happening.

Summary

5.26 In reflecting on the range of efforts to connect teaching development and pupil outcomes it may be helpful to distinguish three levels: the national system, the school and the individual teacher, see Table 5 for examples.

Table 5. Connecting teacher development and pupil outcomes: some examples

System wide	School level	Individual teachers
Finnish ITE is provided at Master's level and includes a strong research element across all teacher education routes. Teacher education students are introduced to research at an early stage and are required to complete a series of small-scale pedagogical inquiries before submitting a research-based Masters thesis (Ostinelli, 2009). The research-based approach has been in place for 30 years. A study by Jyrhämä et al (2008) suggests that pupils appreciate the approach by student teachers who have been working in classrooms prior to doing teacher education and favour a more experiential approach. Tryggvason (2009) asks why Finnish teacher education is successful and finds on the basis of interviews that 'Finnish teacher educators aim to educate reflective and exploring teachers by using a variety of methods in their own pedagogy' (p.369). Hargreaveset al, (2007) suggest that detailed attention to learning (curriculum and pedagogy) is a basis for high performance, rather than prioritizing performance through high-stakes testing.	In Alberta, Canada, there is a well-established school improvement scheme that is promoted across the whole state but is implemented at a school level (Taylor et al, 2006). The emphasis is on community engagement and so although each school develops its own approach, every school is encouraged to engage parents, students and other groups in improvement processes. The state provides a wide range of support and monitoring services for schools, including, for example, reviews of professional development for teachers 11. A report commissioned in Alberta concludes : The importance of professional development that is sustainable and connected to meaningful, locally based contexts reiterates the strong connection developing in Alberta between school improvement initiatives, research-based inquiries such as collaborative and action research and professional development programs. (In-Praxis, 2006:46)	Many of the Teachers for a New Era ( TNE) 12 schemes in the USA have sought to follow individual students through into their school postings and ascertain whether the TNE approach to their preparation leads to improved outcomes for students. The most sustained work has been done at Boston College under the leadership of Marilyn Cochran-Smith (Ludlow et al 2008). Through creating 'a culture of evidence and enquiry', the Boston team has been seeking to identify linkages between students' experience, learning and success as a teacher, by following students from their pre-service program through to their first teaching post. As yet, little has been published by way of findings, but this will be important material to review when it is available. Similar work is being undertaken as part of the STNE project at Aberdeen and has been devised in discussion with Boston College colleagues.

Implications for Scottish teacher education

Although there is little hard evidence to demonstrate the connections between teaching quality, teacher education and pupil outcomes, there is widespread professional agreement that they are positively related. These linkages are under-researched, not least in the Scottish context. The research reviewed suggests that consideration should be given to the national, school-based and individual levels of provision. It is likely that successful developments will require intervention/support at all three levels.